NOAA's Cooperative Observers (COOP) network reports precipitation data that have been the most important data source in describing long term climate change and variability. Unfortunately, when it comes to analyzing precipitation patterns such as frequency and intensities, as in response to the global warming scenario, daily COOP precipitation data are not sufficient. Dedicated volunteer observers are not error-free in making and reporting observation time and amount. Recent reprocessing of 15 years of historical Hydrometeorological Automated Data System (HADS) data at NOAA's National Climatic Data Center (NCDC) provided an opportunity to compare daily COOP data with HADS hourly precipitation data. As part of a rigorous quality assurance of both datasets, we have developed an optimization method that minimizes the differences of daily COOP and nearby time-aggregated HADS precipitations subject to co-raining condition. The co-raining condition is a constraint that both COOP and HADS must report precipitation so that objective function can be established. The minimizer of the objective function is time of day. This method has been tested on heavy rain cases in Western North Carolina, where remnants of hurricanes Frances and Ivan caused many landslides in September 2004. The spatial distribution of relative observation times with respect to COOP's observation time showed consistency with direction of rain producing system, namely, anisotropic nature of precipitations. In addition, possible errors of COOP observation times are detected.